Several reports have shown increases in the prevalence of non-specific symptoms in the general population. Research in the military tends to focus on comparisons between deployed and non-deployed personnel and does not examine trends over time. 4,257 and 4,295 male participants of the Gulf war and Iraq war studies not deployed to either of these wars were randomly sampled and surveyed in 1997/1998 and 2004/2006 in two independent cross-sectional studies. Information was collected on 50 symptoms and the General Health Questionnaire (GHQ-12). Factor analysis was performed to identify an underlying pattern of symptom dimensions, and multivariate regressions were carried out to examine changes in symptom dimensions between the two surveys and the possible role of psychological morbidity. Factor analysis identified a robust pattern of eight symptom dimensions. An increase in the prevalence of symptoms was evident across all symptom dimensions. Adjustment for demographic and service characteristics revealed increases in the odds of scoring highly on symptom dimensions, varying from odds ratios 1.57, 95% CI 1.36-1.81 (cardio-respiratory dimension) to 2.24, 95% CI 1.93-2.60 (fatigue dimension). Unexpectedly, increases were even greater when adjusting for psychological morbidity. There is clear evidence of an increase in the reporting of non-specific symptoms over a 7 year period in the UK Armed Forces. It suggests that the threshold for reporting symptoms has decreased and cannot be explained by psychological distress. The possible implication of this trend for medical practice in the wider population deserves close scrutiny.