Abstract The cadang-cadang associated RNA species, ccRNA-1 and ccRNA-2, were recovered from diseased coconut palm leaf using a range of procedures. These included varying the salt molarity, type of buffer, and reducing agent in the extraction medium, using either polyethylene glycol (PEG) or ammonium sulphate to precipitate the RNA-containing component from tissue extracts, and using three alternative procedures to extract nucleic acid from the PEG-precipitated material. The view that ccRNA-1 is uniquely associated with the disease was supported by showing that the similarly migrating RNA, which sometimes could be recovered from normal palms, differed structurally from ccRNA-1. Kinetics of digestion with ribonuclease A and the single-strand specific S1 endonuclease, and thermal denaturation studies, indicated that the ccRNA-1 molecule has both double- and single-stranded regions. Because disruption of base-pairing by melting or hot formamide treatment failed to affect the apparent molecular weight of ccRNA-1, it is concluded that ccRNA-1 has a looped or “hairpin” secondary structure, stabilized by base-pairing. Its buoyant density (1.62 g/cm3) falls between the expected values for a single-stranded and a double-stranded viral RNA. A comparative study with potato spindle tuber viroid (PSTV) showed that the S1 nuclease digestion kinetics of ccRNA-1 and PSTV were similar. In denaturing formamide polyacrylamide gels, the molecular weight of ccRNA-1 was below that of PSTV and was estimated to be 63,000–73,000. ccRNA-1 is soluble in 2 M LiCl, is able to penetrate 20% polyacrylamide gels, and its thermal denaturation profile responds to Mg2+ like that of citrus exocortis viroid RNA. The structure of ccRNA-1 therefore appears to resemble closely that of viroid RNA. Unlike the known viroids, ccRNA-1 is not sedimented from tissue extracts by low or high speed centrifugation, nor has its infectivity been demonstrated. Some variations in the ccRNA banding patterns in gels are described. The occasional appearance of duplex ccRNA-1 is attributed to the existence of two ccRNA-1 molecules which have the same number of base pairs, but which differ in having larger or smaller amounts of single-stranded RNA associated with them.