Loch Ness Marathon 2026: Race Report, Splits, and What Went Wrong at 28km

Oct 4

6 min read

Finish time: 4:39:53. Projected finish at 27km: around 3:50.

That gap is the whole story, so let's start there and work backwards.

The build

I ran Brighton in April, took a month off, and started running again in June with no plan whatsoever. No structure, no targets, no progression. Just getting out and remembering why I liked it.

That lasted two months and I don't regret a single run of it. It wasn't optimal and it wasn't meant to be.

In August I looked at the data properly instead of going on how I felt, and the picture was uncomfortable. Almost all of my running was sitting in heart rate Zone 3. The moderate middle. Easy runs slightly too quick to build an aerobic base, hard runs not quite hard enough to be a genuine stimulus. Everything piling into one grey band that produces fatigue without producing much adaptation.

So the eight week block that started on 3 August had one job before anything else: separate the easy from the hard.

Easy runs at 6:00 to 6:30 per kilometre with heart rate under 152, which felt embarrassingly slow for the first fortnight. Tempo at 4:45 to 5:00, comfortably hard and controlled. Long runs at easy pace with marathon pace segments built into the back end, climbing from 18k.

Marathon pace target was 5:13, which puts you on 3:40.

Three gym sessions a week alongside it, capped at 90 minutes, two squat days and a deadlift day, working at roughly 85 to 90 percent of training max on the big lifts. Heavy and brief, so strength held without eating into running recovery.

I'll break the full method down in a separate post. This one is about the race.

Two things that didn't go to plan

I raced the Northampton Half when I shouldn't have. It was meant to be a hard training run. I went too hard, tried to add distance afterwards, and stopped at 3.6k. That cost me the 32k peak run in week six, which meant my longest run of the entire block was 28k rather than 32k.

My knee flared ten days out. I stopped running entirely seven days before the race and never completed the second taper week.

Both of those matter later.

The course

Loch Ness is a net downhill point to point with 312m of elevation gain and a high point of 311m at the start.

That sounds generous. It isn't, particularly.

The first 16km drops roughly 280m. Downhill running loads your quads eccentrically, and the damage accumulates whether you feel it at the time or not. Go out hard on a descent and the bill arrives at 30k when you've got nothing left to pay it with.

Then there's a climb between roughly 30k and 36k, right at the point where you're least equipped for it, before a final descent into Inverness.

Phase one: kilometres 1 to 16

Average pace 5:21/km. Average heart rate 145bpm. Elapsed 1:25:29.

The whole plan for this section was to not bank time.

The temptation on a descent like that is enormous. Everything feels easy, the pace reads fast for almost no effort, and you convince yourself you're putting minutes in the bank for later. You aren't. You're putting damage into your quads.

So I held it steady. Heart rate in the 120s and 130s for the opening kilometres, drifting into the 140s as it levelled. Kilometre 9 has a 38m rise and the pace reflects that honestly at 5:47, which is fine. Kilometre 16 at 5:52 was the only genuinely slow one in the phase.

The one I'd query is kilometre 11 at 4:55, which was faster than it needed to be. Forty three metres of descent carrying me rather than me controlling it.

Phase two: kilometres 17 to 27

Average pace 5:16/km. Average heart rate 159bpm. Elapsed 58:01.

This is where the race actually happened.

Course flattens out, and flat is where I'm strongest, so this is where I chose to spend what I had. Kilometres 20 through 25 ran 5:12, 5:12, 5:08, 5:06, 5:05, 5:09. Heart rate climbed steadily from 152 to 163 and sat there.

Nothing forced. No surging. Just a sustainable effort on terrain that suited me, with the climb at 30k in mind and something deliberately held back for it.

Through 27km I was averaging 5:19 per kilometre.

Hold that to the line and you finish in 3:47:35. Allow for the climb between 30 and 36, and realistically you're looking at somewhere around 3:50.

That's eight minutes inside my personal best of 3:58:11. It's also ten minutes outside the 3:40 I trained for, so I'm not going to pretend the target was still live.

Phase three: kilometres 28 to the finish

Average pace 8:39/km. Average heart rate 137bpm.

At 28km the arch of my left foot went.

Sharp pain radiating from the arch into the joint of my big toe with every footstrike. Not a build up, not a warning. It arrived.

I'd never had a problem with that foot in my life. Not in training, not in two previous marathons, not once.

Kilometre 28 was 5:45. Kilometre 29 was 5:55. Then 30 was 7:07 and 31 was 9:03, and from there it was walking with running attempts in between. There were points where I couldn't manage ten strides before having to stop. Kilometres 36, 37, 38 and 40 came in between 11:11 and 12:21. Before managing to pick up the pace a with a steady jog towards the end.

The knee and the ankle held

Worth saying clearly, because I spent eight weeks worrying about the wrong things.

I nursed an ankle niggle through the entire block. Ten days out my knee flared badly enough that I stopped running completely for the final week and skipped my second taper week. I went to the start line convinced one of those two was going to end my race.

Neither of them gave me a single problem. Instead, it was something else that turned out to stop me.

Was the missed peak run the cause?

I don't know, and I'm not going to claim I do.

What I can say is that my longest run of the block was 28km, and the thing broke at 28km. That's hard to ignore, and the honest version is that I raced a half marathon I shouldn't have raced and lost my peak long run because of it.

Whether a 32k run four weeks earlier would have conditioned that arch to hold, or whether this was always going to happen on that day, I genuinely can't tell you. However, from a lot comments on my race video - a lot of people were saying they had the same problems with their feet, knees, hips, or legs overall. The hills destroyed them.

Inverness

The last few kilometres into the city were the best part of the day, and I was barely running.

People shouting your name off your race number. High fives from strangers. Enough noise that walking started to feel like letting them down. I ran what I could and shuffled the rest, and that crowd dragged me to the line far more than any training block did.

That part of a race is impossible to simulate and impossible to overstate.

The numbers

Finish: 4:39:53 · 42.82km · 6:32/km average · 144bpm average · 312m elevation gain · 3,418 calories

My slowest of three marathons. Strava calls it my third fastest, which is a generous way to describe a field of three.

Full splits

Km	Pace	Elev	HR
1	5:25	-43	127
2	5:11	-48	123
3	5:29	-5	134
4	5:46	-16	137
5	5:13	-9	140
6	5:18	-21	144
7	5:16	-8	144
8	5:13	-14	146
9	5:47	+38	159
10	5:04	-37	151
11	4:55	-43	155
12	5:16	-11	153
13	5:24	-6	154
14	5:03	-26	156
15	5:17	-16	149
16	5:52	-23	147
17	5:27	+5	152
18	5:27	-1	156
19	5:25	+1	158
20	5:12	+2	157
21	5:12	0	158
22	5:08	+6	161
23	5:06	-5	161
24	5:05	+1	159
25	5:09	-2	163
26	5:24	+2	163
27	5:26	-4	161
28	5:45	-7	168
29	5:55	+7	166
30	7:07	+16	154
31	9:03	+25	141
32	8:38	+20	141
33	6:55	-13	142
34	7:24	-25	140
35	6:40	-15	148
36	11:30	+20	123
37	11:11	-19	111
38	12:21	-12	108
39	7:47	-7	138
40	12:21	-5	109
41	8:25	-8	127
42	8:49	0	131
42.8	7:39	-1	142

What I'd take from it

You can train for everything except the thing that actually happens.

The block worked. Separating easy from hard gave me an aerobic base I didn't have in June, and the proof is in phase two: eleven kilometres at 5:16 with heart rate holding steady, feeling comfortable, 27km into a marathon. That doesn't happen on grey zone training.

The strength work held up too. My legs were fine. They were still fine at 38km while I was walking.

What beat me was a problem that was completely out of my control. Whilst it can be frustrating that it completely ruined my run. I was grateful to be able to cross over that line with the challenges on the day.

Next up I'm publishing the full Marathon Blueprint: the exact structure behind my prep, the run types and why each one exists, the gym programming, and how to build it for your own race. Free to download.

A lot of you have asked me how to train for a marathon and how to incorporate gym work into it so you're managing your performance and recovery safely and optimally. This will give you all you need to know.

ANDREW CAINE PT