For interlaboratory second opinions requested by patients or clinicians, whether all or selected slides should be sent is unknown. We compared the disagreement rate for requested second opinions for 4 years of sending selected slides with the rate for 1 year of sending all slides. Disagreements were identified in 81 (13.6%) of 596 cases (468 selected slides, 128 all slides). The disagreement rate was less when sending selected slides (n = 58 [12%]) vs all slides (n = 23 [18%]; P = .03). Only 5 cases were identified with disagreement related to whether all or selected slides were sent (selected slides, 1 case [slide of interest inadvertently not chosen]; all slides, 4 cases; P = .03). When all slides were sent (disagreement in number of positive nodes, 2 cases; failure of second laboratory to identify microinvasive ductal carcinoma when present in 1 of 10 core biopsy slides, and ductal carcinoma in situ [DCIS] adjacent to invasive ductal carcinoma when DCIS present on only 1 of 5 slides [2/81 (2%)], 1 case each). Sending all slides is associated with a higher disagreement rate than sending only selected slides, but this disagreement comes from correcting and introducing error.