We analysed by age and sex the inter-laboratory variation in submission rates and positivity rates of urine samples from primary care that were submitted to seven microbiology laboratories within the South West of England. There was an almost twofold difference between the lowest and highest submission and positivity rates across all sex and age groups. This could not be accounted for by differences in population, age mix or numbers of nursing home beds. Increased submission in children will increase diagnosis of urinary tract infection (UTI). In contrast, over-investigation in other age groups increases laboratory and primary care costs and, in the asymptomatic elderly, may lead to unnecessary antibiotic treatment. Continued education is needed in primary care to improve the management of urinary symptoms. Laboratory-based studies of UTI are likely to underestimate the true incidence of this infection substantially. Standardised protocols for urine specimen submission are essential for practices and laboratories participating in surveillance of UTI and antibiotic resistance. In our discussion we offer seven recommendations for improving UTI investigation in general practice.