Steganography is the science of hiding data in otherwise plain text or images. Fractal image compression is a lossy method for compressing a graphical image. The compression involves finding a representation of an image in terms of a set of fractals. A fractal is a mathematical formula for describing an image or sub-image. The new and novel steganographic method detailed in this thesis involves producing a new image that is visually and statistically similar to the original image. The produced new image will have steganographic data hidden in it. The method exploits the fact that image data isn’t sensitive to noise. The algorithm selects regions of the image that are visually least sensitive to modification. Fractal image compression theory is used to construct fractal transforms which when applied to other regions of the image produce slight modifications in these least sensitive areas. A modified area is visually similar to the original region. The modified area is also statistically similar to surrounding areas. Hiding the steganographic data involves producing a new region in the manner described for each bit of steganographic data. The security of the method was tested under various conditions. The results show that the method is a very secure means of hiding and retrieving steganographic information. An Investigation was made in a series of techniques which allow steganographic information to be hidden in and retrieved from multimedia documents in a robust manner.