Whereas nonsteroidal antiinflammatory drugs augment spinal morphine on Day 1, the analgesia gained by simply combining these drugs with conventional "on request" oral regimens on Day 2 is less clear. In this trial, we randomized 80 women undergoing elective cesarean delivery with spinal morphine (0.2 mg) to receive naproxen (500 mg) or placebo every 12 h after surgery. Both groups received conventional therapy with acetaminophen with codeine (on request) and rescue IM opioids. Incision pain on sitting (IPS), incision pain at rest, uterine cramping, and gas pain were evaluated with visual analog scales (0 100). Worst,interval pain (0-10), analgesic use, and side effects were measured over 72 h. At 36 h (primary outcome), naproxen use was associated with reductions in IPS (38.2 +/- 26.0 versus 51.4 +/- 25.7, P = 0.05), incision pain at rest, uterine cramping, and worst interval pain scores. Clinically modest, statistically significant reductions in IPS (P = 0.0001) and opioid use were found over time (P < 0.01). Reductions in the incidence of inadequate analgesia and improvements in overall pain relief (P = 0.0006) on Day I did not persist on Day 2 (overall pain relief, P = 0.057, inadequate analgesia, 24% naproxen versus 27% controls; P = 1.00). The addition of regular doses of naproxen to conventional oral pain therapy after cesarean delivery leads to reductions in IPS at 36 h and pain over Day 2 but does not reduce the incidence of inadequate analgesia.