You may very well have a combination of mitochondrial DNA and genomic DNA in your sample. However, you can do a little bit of research into isolation methods to ensure that your cell/tissue sample does not contain any mitochondria before you isolate DNA. Barring this, there are a few ways you can test your sample for the extent of contamination, and the very presence of contamination itself. Once you are aware of the contribution mitochondrial DNA makes in your sample, you can learn to subtract this out in your analysis.
Each mitochondria contains many copies of a single, circular chromosome of approximately 16 kb. The measley 16 kb of code would be easily buried within a larger genomic DNA sample. However, there are many mitochondria in a cell, as compared to only one nucleus.
Why not start by eliminating the possibility of mitochondria in your sample? How is your DNA isolated and prepared for gel loading? If you homogenize and centrifuge your cells prior to DNA isolation, then the answer is fairly easily arrived at--mitochondria and nuclei fractionate at different centrifugal speeds. However, if you use alkaline lysis, then what you pellet is probably a little bit of both kinds of DNA. If the DNA is then sheared, then both mitochondrial and genomic DNA should be fragemented to the same size and subsequently would run together in the band. A simple experiment would be to isolate DNA two ways. From the same amount of starting material, use one method to separate mitochondrial DNA from nuclear DNA and then isolate the DNA from the two separately, and another to indiscriminately isolate all DNA. Run the samples on a gel (nuclear DNA, mitochondrial DNA, total DNA) side by side and then stain it. You will get an idea of the relative size and amount of mitochondrial DNA in a given sample. (Can it even be seen on your gel? How intense is the band? Is there anything left over in the wells?)
There is generally a lot of genomic DNA left in a well of a standard gel if the loaded sample is very concentrated and/or improperly fragmented. DNA will coil in on itself to varying degrees, and the least coiled samples will run "larger" on the gel. These may seem so large that they don't even fit through the pores in your gel, thus staying behind in the well. Coiled portions will run through at different speeds, depending on how tightly they are coiled. (More coiled means "smaller" and that DNA will run faster through the gel.) Varying the concentration of your gel might help you leave less behind in the well.
If you cannot practically change your methods of isolation, you may have to probe your sample in some way for the presence of possible mitochondrial DNA. You might consider using gene-specific PCR primers and your DNA sample to try and isolate a mitochondria-specific gene. Real-Time PCR is very useful for this purpose, and there are already primer sets for mitochondrial genes available. Try these two websites for more details:
Lastly, you might consider probing your DNA sample for SNPs originating from the mother, as mitochondrial DNA is inherited through the mother only.