how did you check "specificity" of your primers? are you sure there are no unintended targets to be amplified (Primer-blast)?

if the answer is NO, maybe you should try a gradient PCR to find the optimal annealing temperature for your primers. if you cannot do a gradient PCR, you will have to try different annealing temperatures in independent PCRs. I would suggest to increase the temperature. in alternative, work on Magnesium concentration to make the reaction more specific. less Mg2+  in the reaction makes more specific the amplification by DNA polymerase.

It is possible that you use to low temperature for annealing! First one, check the melting temperature of your primers. Second, use this temperature. Third, PCR still unspecific, use step by step (0.5 °C) a higher annealing temperature.

The higher the temperature the higher the specificity.

i agree that raising the annealing temperature is the way to go or add DMSO up to 5% which has a similar effect. 2 questions though...are you using single species dna template ( eg human) or something like soil samples where there may be other species present. What are the band sizes ...is one the correct size but is the other a good clean pand or a small diffuse band which could be primer dimer. Sometimes you can get unusual amolification if you use too much template ...how much are you using?

all boils down to our (wrong) concept of primer design,  priming event  at particular conditions is complex when number of unintended but similar primer binding sites exist

Thanks for all , I have extract the DNA from pure culture of Aspergillus niger and i would to confirm the  species. The primer that i use i have take it from a publish article and i use the same parameter for PCR reaction, Anneling temprture is 55C.

Dear Oadi,

as a precaution, always blast the primer sequences to check there are no mismatches to your intended target, even when you get them from the literature. Regarding the annealing temperature, bear in mind that there could be differences in temperature ramps etc from one PCR machine to another, so some optimization may be necessary. 

First, blast your primers and watch for their parameters (I use oligocalc).

If it is OK, optimize your PCR conditions by doing gradient temperature (X axis) and Mg2+ gradient (Y axis) in a PCR assay. Use small quantities of DNA (10 ng / well). Try different PCR enzymes/kits.

If it doesn't work, re-order your primer. It's possible there was an error during the production (I had the case...). Also, are you sure from the purity of your aspergillus culture ? If the problem remains, you should purify the different DNA bands obtained and sequence them. You could also try universal 16S primers published by Muyzer for DGGE. By blasting the sequences obtained with these primers you would be able to confirm species.

Good answers given already, go through all suggested options carefully and you should come right, good luck (yes there is an element of luck in the lab!)