The contribution of extracellular components to the measurement of pHMRS of a variety of rat tumours (nitrosomethyl urea induced mammary tumours, GH3 prolactinomas, Hepatoma 9618a, UA hepatomas and Walker sarcomas) has been assessed. Acid extractable P(i) was between 2.6 and 12.5 mumol/G wet wt depending on tumour type, and of this 53 +/- 4.8% (mean +/- SEM) was MRS-visible. The P(i) content of tumour exudate was 2-3 mM, of interstitial fluid (sampled from a micropore chamber incorporated within a tumour) 1.7 mM, and of blood plasma 1.95 mM. The mean extracellular volumes of the tumours, measured by distribution of 3H2O and [14C]inulin, were 49-55% depending on tumour type and were at least twice that found in normal liver. Calculations suggested that for most tumours with an extracellular volume not exceeding 55%, at least 65% of the P(i)(MRS) signal was derived from intracellular P(i), and thus that pH(MRS) is a measure of pHi. For each tumour type, pHMRS was measured both in 'pulse-acquire' mode at 1.9 T which may include signals from surrounding tissue, and in localized mode at 4.7 T where the signal came uniquely from tumour tissue. The steady state pHMRS was either neutral or on the alkaline side of neutrality (pH range 7.04-7.37). Raised lactate content and decreased buffering capacity (compared to normal tissues) accompanied these neutral to alkaline pH values.(ABSTRACT TRUNCATED AT 250 WORDS)