BACKGROUND Treatment of cancer with chemotherapy is becoming increasingly more effective but is associated with short and long-term side effects. Oral side effects remain a major source of illness despite the use of a variety of agents to prevent them. OBJECTIVE To evaluate the effectiveness of oral (and topical) prophylactic agents for oral mucositis and oral candidiasis in patients with cancer (excluding head and neck cancer), compared with placebo or no treatment. METHODS Computerised MEDLINE, EMBASE, CINAHL, CANCERLIT, the Cochrane Controlled Trials Register and the Cochrane Oral Health Group Specialist Register search up to July 1999. Reference lists from relevant articles were scanned and the authors of eligible studies were contacted to identify trials and obtain additional information. METHODS Studies were selected if they met the following criteria: design - random or quasi-random allocation of participants; participants - anyone with cancer receiving chemotherapy (excluding head and neck cancer); interventions - prophylactic agents prescribed to reduce oral conditions arising from cancer or its treatment; outcomes - mucositis and oral candidiasis. METHODS Information regarding methods, participants, interventions and outcome measures and results were independently extracted, in duplicate, by two reviewers (JC & HW). Specialist advice was sought to categorise interventions. Authors were contacted for details of randomisation and withdrawals and a quality assessment was carried out using the Jadad criteria (Jadad 1998). The Cochrane Oral Health Group statistical guidelines were followed and relative risk values calculated using random effects models where significant heterogeneity was detected (P < 0.1). RESULTS Thirty-eight reports of trials were initially included. Two were duplicate reports and nine were excluded as there was no useable information. Of the 27 useable studies 14 had data for mucositis comprising 945 randomised patients and 15 included data for oral candidiasis with 1164 randomised patients. Of the eight prophylactic agents used for mucositis only one, ice chips, was effective (Relative risk 0.57, 95% CI 0.43 to 0.77, chi-square for heterogeneity = 0.26 (df = 1), p = 0.61). The NNT to prevent one extra case of mucositis over the baseline incidence using ice chips was 4 (95%CI: 3 to 7). The NNT for when the baseline incidence of mucositis in the population ranges from 50% to 80% are 5 to 4 respectively. There is evidence that antifungal agents which are partially or fully absorbed from the gastrointestinal tract prevent oral candidiasis and that the partially absorbed agents may be more effective than the fully absorbed agents. The RR for partially absorbed agents was 0.13 (95% CI 0.06 to 0.27, chi-square for heterogeneity = 5.3 (df = 3), P = 0. 15). The NNT to prevent one extra case of oral candidiasis over the baseline incidence using partially absorbed drugs was 3 (95% CI: 3 to 5). The NNT for when the baseline incidence of oral candidiasis in the population ranges from 30% to 70% are 4 to 2 respectively. The general reporting of RCT's was poor however the median Jadad score was acceptable and improved further when the authors provided additional information. The sensitivity analysis confirmed the findings for oral candidiasis. CONCLUSIONS There is some evidence that ice chips prevent mucositis. None of the other prophylactic agents included in this review prevented mucositis. There is evidence that prophylactic use of antifungal agents which are absorbed or partially absorbed from the gastrointestinal tract reduce the clinical signs of oral candidiasis, and the partially absorbed drugs may be more effective. Future trials in this area should address the link between oral and general health including outcomes relevant to the patient. Collaboration between medical and dental teams is indicated.