Effective treatment and prevention of urolithiasis depends on accurate determination of the chemical nature of the uroliths. A widely used qualitative chemical procedure was compared with quantitative crystallographic analysis of 272 canine uroliths. Agreement between the 2 methods was 78%. Qualitative analysis failed to detect 62% of calcium-containing uroliths and 83% of carbonate apatite uroliths. Qualitative analysis gave false-positive results for urates in 55% of cystine uroliths. Mixed uroliths comprising 6% of the total could not be classified without quantitative analysis. Silicate, cystine, and urate uroliths generally were of pure composition. Crystallographic analysis indicated the following distribution of major types: struvite, 69%; calcium oxalate, 10%; urate, 7%; silicate, 3.5%; cystine, 3.2%; calcium phosphate, 1%; and mixed, 6%. Among dogs with struvite uroliths, 66% had positive results of bacterial culturing from the urinary bladder. Six breeds (Miniature Schnauzer, Welsh Corgi, Lhasa Apso, Yorkshire Terrier, Pekingese, and Pug) had a significantly higher risk for urolithiasis, compared with other breeds. The German Shepherd Dog had a significantly lowered risk, compared with other breeds. Two breeds had significant relationship to a specific type of urolith: Miniature Schnauzer for oxalate, and Dalmatian for urate (P less than 0.001). It was concluded that quantitative analysis, using crystallography, was superior for the detection of calcium oxalate, carbonate apatite, cystine, urate, and mixed uroliths.