Double-blind randomized controlled trials in single subjects (N of 1 RCTs) have demonstrated a beneficial symptomatic effect of cimetidine in reflux- or ulcer-like non-ulcer dyspepsia (NUD). However, spontaneous fluctuations in symptoms reduce the validity of such trials when performed as continuous trials with fixed dosages. This study was carried out to identify individual responders to cimetidine in NUD, peptic ulcer disease, and oesophagitis and to confirm the beneficial average effect of cimetidine in these clinical entities. We evaluated N of 1 multi-crossover trial designs, which compare the effects of single doses of cimetidine and placebo taken on-demand for symptomatic relief. Each trial consisted of six cimetidine (400 mg or 800 mg) and six placebo tablets randomized in successive pairs. The symptomatic effect of each tablet was measured 1/2-6 h after the intake. Outcomes were assessed by individual p values and confidence intervals. A minimal clinically important difference was defined, to assess the clinical significance as demonstrated by the confidence intervals. Thirteen of 25 patients (52%) with reflux- and ulcer-like NUD obtained individual p values below 0.20. Similarly, 7 of 9 patients (78%) with oesophagitis and 6 of 12 patients (50%) with peptic ulcer obtained such p values. On the basis of the 80% confidence intervals the corresponding numbers of subjects with clinically significant effect were six (NUD), three, and three. The combined data showed a significantly better effect of cimetidine than of placebo (p less than 0.0001) in each of the three diagnostic groups studied. Cimetidine taken on-demand may have a rapid symptom-relieving effect in dyspepsia.(ABSTRACT TRUNCATED AT 250 WORDS)