1. Acutely, opioids inhibit oxytocin secretion. To study the responses of oxytocin neurones during chronic opioid exposure, forty-five lactating rats were infused continuously from a subcutaneous osmotically driven mini-pump via a lateral cerebral ventricle with morphine sulphate solution from day 2 post-partum for 5-7 days; the infusion rate was increased 2- or 2.5-fold each 40 h from 10 micrograms/h initially up to 50 micrograms/h; controls were infused with vehicle (1 microliter/h, twenty-eight rats) or were untreated (eight rats). 2. Maternal behaviour was disrupted in 27% of the morphine-treated rats; in rats that remained maternal morphine did not affect body weight or water intake but increased rectal temperature by 0.82 +/- 0.14 degrees C (mean +/- S.E.M.) across the first 4 days. 3. Weight gain of the litters of maternal morphine-treated rats was reduced by 32% during 7 days, predominantly in the first day of treatment when milk transfer was also reduced. Observation of pup behaviour during suckling showed decreased frequency of milk ejections on only the second day of morphine treatment. Plasma concentration of prolactin after 6 days was similar in maternal morphine-treated and control rats, but reduced by 90% in non-maternal morphine-treated rats, indicating normal control of prolactin secretion by suckling in morphine-treated rats. 4. Oxytocin and vasopressin contents, measured by radioimmunoassay, in the supraoptic and paraventricular nuclei and in the neurohypophysis were similar between fourteen maternal morphine-treated, twelve vehicle-treated and eight untreated lactating rats; thus exposure to morphine did not involve increased production and storage of oxytocin. 5. Distribution of [3H]morphine infused intracerebroventricularly into six virgin female rats for 6 days was measured by scintillation counting of tissue extracts. Morphine concentration in the hypothalamus and neurohypophysis was 2.7 and 12.8 micrograms/g, respectively, and in blood plasma 0.75 micrograms/g. Tolerance was not due to failure of morphine infusion. In addition, naloxone (5 mg/kg s.c.) provoked typical withdrawal reactions ('wet dog' shakes, defaecation, burrowing) in lactating rats infused with morphine for 5 days. 6. Pups were suckled onto seven maternal morphine-infused and five vehicle-infused rats anaesthetized with urethane for recording of intramammary and arterial blood pressures after treatment for 5 days. The incidence and pattern of milk ejections, and mammary gland sensitivity to oxytocin were similar in the two groups.(ABSTRACT TRUNCATED AT 400 WORDS)