THE induction of sarcomas at the site of injection of iron-dextran, of otlher preparations of iron, and of other metal-carbohydrate complexes in rats, mice, hamsters and rabbits is well documented (Richmond, 1957(Richmond, , 1959(Richmond, , 1960Haddow and Horning, 1960; Golberg, Martin and Smith, 1960 ; Lundin, 1961 ;Haddow. Dukes and Mitchley, 1961;Haddow, Roe and Mitchley, 1964;Haddow and Roe, 1964; Roe, Haddow, Dukes and Mitchley, 1964) and has been considered in recent review articles (Roe and Lancaster, 1964;Roe, 1965).The iron sorbitol-citric acid complex marketed under the name of " Jectofer' has hitherto received less attention from the point of view of carcinogenicitv. Lundin (1961) reported an experiment in which male or female Sprague-Dawley rats were given repeated injections of Jectofer. The rats were injected twice weekly with Jectofer at a rate of 0 05 ml. per injection for each 50 g. of body weight. In this way each rat received on average a total of 255 mg. of iron over the 4 month period. All injections were made intramuscularly into the right thigh. Thirty-eight animals were kept under observation from the 38th to the 68th week of the experiment. Amongst these, one rat developed a benign fibroma at the injection site. Comparable groups of rats injected similarly with iroindextran (Imferon) or Ferrigen (a high molecular weight iron-carbohydrate complex marketed by Astra Chemicals) developed numerous injection-site tumours, the first appearing around the 40th week. The tumour response was more marked in the case of Imferon than in that of Ferrigen. Rats injected with either Imferon or Ferrigen according to the same schedule, but with twice as much of each compound developed tumours in even higher incidence and after a shorter meain latent interval. Because of systemic toxicity, Lundin was unable to give Jectofer at the higher dose level. Fielding (1962) tested Jectofer for carcinogenicity in mice, including in the same experiment groups treated with iron-dextran (Imferon) and iron-dextrin (Astrafer). All three groups were injected once weekly, subcutaneously in the left flank with the iron preparation such that a dose of 1 mg. iron was given. In the cases of Jectofer and Imferon this entailed a dose volume of 0 02 ml., but in the case of Astrafer, the volume was 0 05 ml. In all groups injections were continued for 28 to 30 weeks and the experiment was terminated after 17 months. No injection-site tumours were seen in 28 Jectofer-treated mice which lived for 12 months or more, whereas 2 out of 17 mice developed tumours in response to iron-dextran and 3 out of 12 to iron-dextrin.The experiment reported in the present paper represents a third attempt to induce tumours by the injection of Jectofer. Once again, as a positive control, animals (rats) were injected with iron-dextran (Imferon). Attention was paid,