Two hundred and fifty six consecutively ambulant diabetic patients attending a Nigerian diabetic clinic for the presence of cheiroarthropathy were examined using standard criteria. Forty eight (19%) had limited joint mobility of the hand, a prevalence higher than the 4% (2/56) observed in a non-diabetic population matched for age and sex. Limited joint mobility was twice as prevalent in the insulin treated (16/50, 32%) than in the non-insulin dependent (32/206, 16%) diabetics. Indices such as age, duration of diabetes, and glycaemic control (as assessed by integrated blood glucose concentration over the previous 12 months and glycated haemoglobin (HbA1c concentrations) were the same in the diabetics with and without limited joint mobility. The prevalence of cataracts (18/48, 38%) and background retinopathy (8/48, 17%) was higher in diabetics with limited joint mobility than in those without (respectively 6-9% and 5-6%); hypertension, peripheral neuropathy, and foot ulcers, however, were about equally common in the two groups of diabetics (with and without limited joint mobility). Nephropathy appeared commoner in diabetic subjects without limited joint mobility. Our results confirm previous observations in Caucasians of an increased prevalence of limited joint mobility in diabetes, especially those receiving insulin treatment, and also showed that limited joint mobility could predict the presence of retinopathy and cataracts in those diabetics. Neuropathy and hypertension were not commoner in our diabetics with limited joint mobility (unlike in the Caucasian population), suggesting that racial factors may underlie the predictive value of limited joint mobility in diabetic microangiopathy.